git-commit-message 0.7.0__tar.gz → 0.8.0__tar.gz

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: git-commit-message
-Version: 0.7.0
+Version: 0.8.0
 Summary: Generate Git commit messages from staged changes using LLM
 Maintainer-email: Mina Her <minacle@live.com>
 License: This is free and unencumbered software released into the public domain.
@@ -45,16 +45,22 @@ Requires-Python: >=3.13
 Description-Content-Type: text/markdown
 Requires-Dist: babel>=2.17.0
 Requires-Dist: google-genai>=1.56.0
+Requires-Dist: ollama>=0.4.0
 Requires-Dist: openai>=2.6.1
 Requires-Dist: tiktoken>=0.12.0
 
 # git-commit-message
 
-Staged changes -> GPT commit message generator.
+Generate a commit message from your staged changes using OpenAI, Google Gemini, or Ollama.
 
 [![asciicast](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN.svg)](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN)
 
-## Install (PyPI)
+## Requirements
+
+- Python 3.13+
+- A Git repo with staged changes (`git add ...`)
+
+## Install
 
 Install the latest released version from PyPI:
 
@@ -78,19 +84,43 @@ Quick check:
 git-commit-message --help
 ```
 
-Set your API key (POSIX sh):
+## Setup
+
+### OpenAI
 
 ```sh
 export OPENAI_API_KEY="sk-..."
 ```
 
-Or for the Google provider:
+### Google Gemini
 
 ```sh
 export GOOGLE_API_KEY="..."
 ```
 
-Note (fish): In fish, set it as follows.
+### Ollama (local models)
+
+1. Install Ollama: https://ollama.ai
+2. Start the server:
+
+```sh
+ollama serve
+```
+
+3. Pull a model:
+
+```sh
+ollama pull mistral
+```
+
+Optional: set defaults:
+
+```sh
+export GIT_COMMIT_MESSAGE_PROVIDER=ollama
+export OLLAMA_MODEL=mistral
+```
+
+Note (fish):
 
 ```fish
 set -x OPENAI_API_KEY "sk-..."
@@ -104,96 +134,116 @@ python -m pip install -e .
 
 ## Usage
 
-- Print commit message only:
+Generate and print a commit message:
 
 ```sh
 git add -A
 git-commit-message "optional extra context about the change"
 ```
 
-- Force single-line subject only:
+Generate a single-line subject only:
 
 ```sh
 git-commit-message --one-line "optional context"
 ```
 
-- Select provider (default: openai):
+Select provider:
 
 ```sh
-git-commit-message --provider openai "optional context"
+# OpenAI (default)
+git-commit-message --provider openai
+
+# Google Gemini (via google-genai)
+git-commit-message --provider google
+
+# Ollama
+git-commit-message --provider ollama
 ```
 
-- Select provider (Google Gemini via google-genai):
+Commit immediately (optionally open editor):
 
 ```sh
-git-commit-message --provider google "optional context"
+git-commit-message --commit "refactor parser for speed"
+git-commit-message --commit --edit "refactor parser for speed"
 ```
 
-- Limit subject length (default 72):
+Limit subject length:
 
 ```sh
-git-commit-message --one-line --max-length 50 "optional context"
+git-commit-message --one-line --max-length 50
 ```
 
-- Chunk long diffs by token budget (0 = single chunk + summary, -1 = disable chunking):
+Chunk/summarise long diffs by token budget:
 
 ```sh
 # force a single summary pass over the whole diff (default)
-git-commit-message --chunk-tokens 0 "optional context"
+git-commit-message --chunk-tokens 0
 
 # chunk the diff into ~4000-token pieces before summarising
-git-commit-message --chunk-tokens 4000 "optional context"
+git-commit-message --chunk-tokens 4000
 
 # disable summarisation and use the legacy one-shot prompt
-git-commit-message --chunk-tokens -1 "optional context"
+git-commit-message --chunk-tokens -1
 ```
 
-- Commit immediately with editor:
+Select output language/locale (IETF language tag):
 
 ```sh
-git-commit-message --commit --edit "refactor parser for speed"
+git-commit-message --language en-US
+git-commit-message --language ko-KR
+git-commit-message --language ja-JP
 ```
 
-- Print debug info (prompt/response + token usage):
+Print debug info:
 
 ```sh
-git-commit-message --debug "optional context"
+git-commit-message --debug
 ```
 
-- Select output language/locale (default: en-GB):
+Configure Ollama host (if running on a different machine):
 
 ```sh
-# American English
-git-commit-message --language en-US "optional context"
+git-commit-message --provider ollama --host http://192.168.1.100:11434
+```
 
-# Korean
-git-commit-message --language ko-KR
+## Options
 
-# Japanese
-git-commit-message --language ja-JP
-```
+- `--provider {openai,google,ollama}`: provider to use (default: `openai`)
+- `--model MODEL`: model override (provider-specific)
+- `--language TAG`: output language/locale (default: `en-GB`)
+- `--one-line`: output subject only
+- `--max-length N`: max subject length (default: 72)
+- `--chunk-tokens N`: token budget per diff chunk (`0` = single summary pass, `-1` disables summarisation)
+- `--debug`: print request/response details
+- `--commit`: run `git commit -m <message>`
+- `--edit`: with `--commit`, open editor for final message
+- `--host URL`: host URL for providers like Ollama (default: `http://localhost:11434`)
+
+## Environment variables
 
-Notes:
+Required:
 
-- The model is instructed to write using the selected language/locale.
-- In multi-line mode, the only allowed label ("Rationale:") is also translated into the target language.
+- `OPENAI_API_KEY`: when provider is `openai`
+- `GOOGLE_API_KEY`: when provider is `google`
 
-Environment:
+Optional:
 
-- `OPENAI_API_KEY`: required when provider is `openai`
-- `GOOGLE_API_KEY`: required when provider is `google`
-- `GIT_COMMIT_MESSAGE_PROVIDER`: optional (default: `openai`). `--provider` overrides this value.
-- `GIT_COMMIT_MESSAGE_MODEL`: optional model override (defaults: `openai` -> `gpt-5-mini`, `google` -> `gemini-2.5-flash`)
-- `OPENAI_MODEL`: optional OpenAI-only model override
-- `GIT_COMMIT_MESSAGE_LANGUAGE`: optional (default: `en-GB`)
-- `GIT_COMMIT_MESSAGE_CHUNK_TOKENS`: optional token budget per diff chunk (default: 0 = single chunk + summary; -1 disables summarisation)
+- `GIT_COMMIT_MESSAGE_PROVIDER`: default provider (`openai` by default). `--provider` overrides this.
+- `GIT_COMMIT_MESSAGE_MODEL`: model override for any provider. `--model` overrides this.
+- `OPENAI_MODEL`: OpenAI-only model override (used if `--model`/`GIT_COMMIT_MESSAGE_MODEL` are not set)
+- `OLLAMA_MODEL`: Ollama-only model override (used if `--model`/`GIT_COMMIT_MESSAGE_MODEL` are not set)
+- `OLLAMA_HOST`: Ollama server URL (default: `http://localhost:11434`)
+- `GIT_COMMIT_MESSAGE_LANGUAGE`: default language/locale (default: `en-GB`)
+- `GIT_COMMIT_MESSAGE_CHUNK_TOKENS`: default chunk token budget (default: `0`)
 
-Notes:
+Default models (if not overridden):
 
-- If token counting fails for your provider while chunking, try `--chunk-tokens 0` (default) or `--chunk-tokens -1`.
+- OpenAI: `gpt-5-mini`
+- Google: `gemini-2.5-flash`
+- Ollama: `gpt-oss:20b`
 
-## AI‑generated code notice
+## AI-generated code notice
 
 Parts of this project were created with assistance from AI tools (e.g. large language models).
-All AI‑assisted contributions were reviewed and adapted by maintainers before inclusion.
+All AI-assisted contributions were reviewed and adapted by maintainers before inclusion.
 If you need provenance for specific changes, please refer to the Git history and commit messages.

git_commit_message-0.8.0/README.md

@@ -0,0 +1,198 @@
+# git-commit-message
+
+Generate a commit message from your staged changes using OpenAI, Google Gemini, or Ollama.
+
+[![asciicast](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN.svg)](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN)
+
+## Requirements
+
+- Python 3.13+
+- A Git repo with staged changes (`git add ...`)
+
+## Install
+
+Install the latest released version from PyPI:
+
+```sh
+# User environment (recommended)
+python -m pip install --user git-commit-message
+
+# Or system/virtualenv as appropriate
+python -m pip install git-commit-message
+
+# Or with pipx for isolated CLI installs
+pipx install git-commit-message
+
+# Upgrade to the newest version
+python -m pip install --upgrade git-commit-message
+```
+
+Quick check:
+
+```sh
+git-commit-message --help
+```
+
+## Setup
+
+### OpenAI
+
+```sh
+export OPENAI_API_KEY="sk-..."
+```
+
+### Google Gemini
+
+```sh
+export GOOGLE_API_KEY="..."
+```
+
+### Ollama (local models)
+
+1. Install Ollama: https://ollama.ai
+2. Start the server:
+
+```sh
+ollama serve
+```
+
+3. Pull a model:
+
+```sh
+ollama pull mistral
+```
+
+Optional: set defaults:
+
+```sh
+export GIT_COMMIT_MESSAGE_PROVIDER=ollama
+export OLLAMA_MODEL=mistral
+```
+
+Note (fish):
+
+```fish
+set -x OPENAI_API_KEY "sk-..."
+```
+
+## Install (editable)
+
+```sh
+python -m pip install -e .
+```
+
+## Usage
+
+Generate and print a commit message:
+
+```sh
+git add -A
+git-commit-message "optional extra context about the change"
+```
+
+Generate a single-line subject only:
+
+```sh
+git-commit-message --one-line "optional context"
+```
+
+Select provider:
+
+```sh
+# OpenAI (default)
+git-commit-message --provider openai
+
+# Google Gemini (via google-genai)
+git-commit-message --provider google
+
+# Ollama
+git-commit-message --provider ollama
+```
+
+Commit immediately (optionally open editor):
+
+```sh
+git-commit-message --commit "refactor parser for speed"
+git-commit-message --commit --edit "refactor parser for speed"
+```
+
+Limit subject length:
+
+```sh
+git-commit-message --one-line --max-length 50
+```
+
+Chunk/summarise long diffs by token budget:
+
+```sh
+# force a single summary pass over the whole diff (default)
+git-commit-message --chunk-tokens 0
+
+# chunk the diff into ~4000-token pieces before summarising
+git-commit-message --chunk-tokens 4000
+
+# disable summarisation and use the legacy one-shot prompt
+git-commit-message --chunk-tokens -1
+```
+
+Select output language/locale (IETF language tag):
+
+```sh
+git-commit-message --language en-US
+git-commit-message --language ko-KR
+git-commit-message --language ja-JP
+```
+
+Print debug info:
+
+```sh
+git-commit-message --debug
+```
+
+Configure Ollama host (if running on a different machine):
+
+```sh
+git-commit-message --provider ollama --host http://192.168.1.100:11434
+```
+
+## Options
+
+- `--provider {openai,google,ollama}`: provider to use (default: `openai`)
+- `--model MODEL`: model override (provider-specific)
+- `--language TAG`: output language/locale (default: `en-GB`)
+- `--one-line`: output subject only
+- `--max-length N`: max subject length (default: 72)
+- `--chunk-tokens N`: token budget per diff chunk (`0` = single summary pass, `-1` disables summarisation)
+- `--debug`: print request/response details
+- `--commit`: run `git commit -m <message>`
+- `--edit`: with `--commit`, open editor for final message
+- `--host URL`: host URL for providers like Ollama (default: `http://localhost:11434`)
+
+## Environment variables
+
+Required:
+
+- `OPENAI_API_KEY`: when provider is `openai`
+- `GOOGLE_API_KEY`: when provider is `google`
+
+Optional:
+
+- `GIT_COMMIT_MESSAGE_PROVIDER`: default provider (`openai` by default). `--provider` overrides this.
+- `GIT_COMMIT_MESSAGE_MODEL`: model override for any provider. `--model` overrides this.
+- `OPENAI_MODEL`: OpenAI-only model override (used if `--model`/`GIT_COMMIT_MESSAGE_MODEL` are not set)
+- `OLLAMA_MODEL`: Ollama-only model override (used if `--model`/`GIT_COMMIT_MESSAGE_MODEL` are not set)
+- `OLLAMA_HOST`: Ollama server URL (default: `http://localhost:11434`)
+- `GIT_COMMIT_MESSAGE_LANGUAGE`: default language/locale (default: `en-GB`)
+- `GIT_COMMIT_MESSAGE_CHUNK_TOKENS`: default chunk token budget (default: `0`)
+
+Default models (if not overridden):
+
+- OpenAI: `gpt-5-mini`
+- Google: `gemini-2.5-flash`
+- Ollama: `gpt-oss:20b`
+
+## AI-generated code notice
+
+Parts of this project were created with assistance from AI tools (e.g. large language models).
+All AI-assisted contributions were reviewed and adapted by maintainers before inclusion.
+If you need provenance for specific changes, please refer to the Git history and commit messages.

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/pyproject.toml

@@ -1,12 +1,13 @@
 [project]
 name = "git-commit-message"
-version = "0.7.0"
+version = "0.8.0"
 description = "Generate Git commit messages from staged changes using LLM"
 readme = "README.md"
 requires-python = ">=3.13"
 dependencies = [
 	"babel>=2.17.0",
 	"google-genai>=1.56.0",
+	"ollama>=0.4.0",
 	"openai>=2.6.1",
 	"tiktoken>=0.12.0",
 ]

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message/_cli.py

@@ -27,6 +27,38 @@ from ._llm import (
 )
 
 
+class CliArgs(Namespace):
+    __slots__ = (
+        "description",
+        "commit",
+        "edit",
+        "provider",
+        "model",
+        "language",
+        "debug",
+        "one_line",
+        "max_length",
+        "chunk_tokens",
+        "host",
+    )
+
+    def __init__(
+        self,
+        /,
+    ) -> None:
+        self.description: str | None = None
+        self.commit: bool = False
+        self.edit: bool = False
+        self.provider: str | None = None
+        self.model: str | None = None
+        self.language: str | None = None
+        self.debug: bool = False
+        self.one_line: bool = False
+        self.max_length: int | None = None
+        self.chunk_tokens: int | None = None
+        self.host: str | None = None
+
+
 def _env_chunk_tokens_default() -> int | None:
     """Return chunk token default from env if valid, else None."""
 
@@ -87,7 +119,7 @@ def _build_parser() -> ArgumentParser:
         "--model",
         default=None,
         help=(
-            "Model name to use. If unspecified, uses GIT_COMMIT_MESSAGE_MODEL or a provider-specific default (openai: gpt-5-mini; google: gemini-2.5-flash)."
+            "Model name to use. If unspecified, uses GIT_COMMIT_MESSAGE_MODEL or a provider-specific default (openai: gpt-5-mini; google: gemini-2.5-flash; ollama: gpt-oss:20b)."
         ),
     )
 
@@ -134,11 +166,21 @@ def _build_parser() -> ArgumentParser:
         ),
     )
 
+    parser.add_argument(
+        "--host",
+        dest="host",
+        default=None,
+        help=(
+            "Host URL for API providers like Ollama (default: http://localhost:11434). "
+            "You may also set OLLAMA_HOST for Ollama."
+        ),
+    )
+
     return parser
 
 
 def _run(
-    args: Namespace,
+    args: CliArgs,
     /,
 ) -> int:
     """Main execution logic.
@@ -177,11 +219,12 @@ def _run(
                 diff_text,
                 hint,
                 args.model,
-                getattr(args, "one_line", False),
-                getattr(args, "max_length", None),
-                getattr(args, "language", None),
+                args.one_line,
+                args.max_length,
+                args.language,
                 chunk_tokens,
-                getattr(args, "provider", None),
+                args.provider,
+                args.host,
             )
             message = result.message
         else:
@@ -189,11 +232,12 @@ def _run(
                 diff_text,
                 hint,
                 args.model,
-                getattr(args, "one_line", False),
-                getattr(args, "max_length", None),
-                getattr(args, "language", None),
+                args.one_line,
+                args.max_length,
+                args.language,
                 chunk_tokens,
-                getattr(args, "provider", None),
+                args.provider,
+                args.host,
             )
     except UnsupportedProviderError as exc:
         print(str(exc), file=stderr)
@@ -203,7 +247,7 @@ def _run(
         return 3
 
     # Option: force single-line message
-    if getattr(args, "one_line", False):
+    if args.one_line:
         # Use the first non-empty line only
         for line in (ln.strip() for ln in message.splitlines()):
             if line:
@@ -218,7 +262,7 @@ def _run(
             print(f"==== {result.provider} Usage ====")
             print(f"provider: {result.provider}")
             print(f"model: {result.model}")
-            print(f"response_id: {getattr(result, 'response_id', '(n/a)')}")
+            print(f"response_id: {result.response_id or '(n/a)'}")
             if result.total_tokens is not None:
                 print(
                     f"tokens: prompt={result.prompt_tokens} completion={result.completion_tokens} total={result.total_tokens}"
@@ -240,7 +284,7 @@ def _run(
         print(f"==== {result.provider} Usage ====")
         print(f"provider: {result.provider}")
         print(f"model: {result.model}")
-        print(f"response_id: {getattr(result, 'response_id', '(n/a)')}")
+        print(f"response_id: {result.response_id or '(n/a)'}")
         if result.total_tokens is not None:
             print(
                 f"tokens: prompt={result.prompt_tokens} completion={result.completion_tokens} total={result.total_tokens}"
@@ -269,7 +313,8 @@ def main() -> None:
     """
 
     parser: Final[ArgumentParser] = _build_parser()
-    args: Namespace = parser.parse_args()
+    args = CliArgs()
+    parser.parse_args(namespace=args)
 
     if args.edit and not args.commit:
         print("'--edit' must be used together with '--commit'.", file=stderr)

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message/_gemini.py

@@ -7,6 +7,7 @@ Provider-agnostic orchestration/prompt logic lives in `_llm.py`.
 from __future__ import annotations
 
 from os import environ
+from typing import ClassVar
 
 from google import genai
 from google.genai import types
@@ -15,7 +16,11 @@ from ._llm import LLMTextResult, LLMUsage
 
 
 class GoogleGenAIProvider:
-    name = "google"
+    __slots__ = (
+        "_client",
+    )
+
+    name: ClassVar[str] = "google"
 
     def __init__(
         self,

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message/_gpt.py

@@ -9,6 +9,8 @@ from __future__ import annotations
 from openai import OpenAI
 from openai.types.responses import Response
 from os import environ
+from typing import ClassVar
+
 from tiktoken import Encoding, encoding_for_model, get_encoding
 from ._llm import LLMTextResult, LLMUsage
 
@@ -24,7 +26,11 @@ def _encoding_for_model(
 
 
 class OpenAIResponsesProvider:
-    name = "openai"
+    __slots__ = (
+        "_client",
+    )
+
+    name: ClassVar[str] = "openai"
 
     def __init__(
         self,

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message/_llm.py

@@ -12,16 +12,19 @@ from __future__ import annotations
 
 from babel import Locale
 from os import environ
-from typing import Final, Protocol
+from typing import ClassVar, Final, Protocol
 
 
 _DEFAULT_PROVIDER: Final[str] = "openai"
 _DEFAULT_MODEL_OPENAI: Final[str] = "gpt-5-mini"
 _DEFAULT_MODEL_GOOGLE: Final[str] = "gemini-2.5-flash"
+_DEFAULT_MODEL_OLLAMA: Final[str] = "gpt-oss:20b"
 _DEFAULT_LANGUAGE: Final[str] = "en-GB"
 
 
 class UnsupportedProviderError(RuntimeError):
+    __slots__ = ()
+
     pass
 
 
@@ -66,7 +69,9 @@ class LLMTextResult:
 
 
 class CommitMessageProvider(Protocol):
-    name: str
+    __slots__ = ()
+
+    name: ClassVar[str]
 
     def generate_text(
         self,
@@ -147,6 +152,9 @@ def _resolve_model(
     if provider_name == "google":
         default_model = _DEFAULT_MODEL_GOOGLE
         provider_model = None
+    elif provider_name == "ollama":
+        default_model = _DEFAULT_MODEL_OLLAMA
+        provider_model = environ.get("OLLAMA_MODEL")
     else:
         default_model = _DEFAULT_MODEL_OPENAI
         provider_model = environ.get("OPENAI_MODEL")
@@ -164,6 +172,8 @@ def _resolve_language(
 def get_provider(
     provider: str | None,
     /,
+    *,
+    host: str | None = None,
 ) -> CommitMessageProvider:
     name = _resolve_provider(provider)
 
@@ -179,8 +189,14 @@ def get_provider(
 
         return GoogleGenAIProvider()
 
+    if name == "ollama":
+        # Local import to avoid import cycles: providers may import shared types from this module.
+        from ._ollama import OllamaProvider
+
+        return OllamaProvider(host=host)
+
     raise UnsupportedProviderError(
-        f"Unsupported provider: {name}. Supported providers: openai, google"
+        f"Unsupported provider: {name}. Supported providers: openai, google, ollama"
     )
 
 
@@ -459,13 +475,14 @@ def generate_commit_message(
     language: str | None = None,
     chunk_tokens: int | None = 0,
     provider: str | None = None,
+    host: str | None = None,
     /,
 ) -> str:
     chosen_provider = _resolve_provider(provider)
     chosen_model = _resolve_model(model, chosen_provider)
     chosen_language = _resolve_language(language)
 
-    llm = get_provider(chosen_provider)
+    llm = get_provider(chosen_provider, host=host)
 
     normalized_chunk_tokens = 0 if chunk_tokens is None else chunk_tokens
 
@@ -513,13 +530,14 @@ def generate_commit_message_with_info(
     language: str | None = None,
     chunk_tokens: int | None = 0,
     provider: str | None = None,
+    host: str | None = None,
     /,
 ) -> CommitMessageResult:
     chosen_provider = _resolve_provider(provider)
     chosen_model = _resolve_model(model, chosen_provider)
     chosen_language = _resolve_language(language)
 
-    llm = get_provider(chosen_provider)
+    llm = get_provider(chosen_provider, host=host)
 
     normalized_chunk_tokens = 0 if chunk_tokens is None else chunk_tokens
 

git_commit_message-0.8.0/src/git_commit_message/_ollama.py

@@ -0,0 +1,122 @@
+"""Ollama provider implementation.
+
+Mirrors the Gemini provider structure: a single provider class that exposes
+`generate_text` and `count_tokens`. Provider-agnostic orchestration lives in
+`_llm.py`.
+"""
+
+from __future__ import annotations
+
+from os import environ
+from typing import ClassVar, Final
+
+from ollama import Client, ResponseError
+from tiktoken import Encoding, get_encoding
+
+from ._llm import LLMTextResult, LLMUsage
+
+
+_DEFAULT_OLLAMA_HOST: Final[str] = "http://localhost:11434"
+
+
+def _resolve_ollama_host(
+    host: str | None,
+    /,
+) -> str:
+    """Resolve the Ollama host URL from arg, env, or default."""
+
+    return host or environ.get("OLLAMA_HOST") or _DEFAULT_OLLAMA_HOST
+
+
+def _get_encoding() -> Encoding:
+    """Get a fallback encoding for token counting."""
+
+    try:
+        return get_encoding("cl100k_base")
+    except Exception:
+        return get_encoding("gpt2")
+
+
+class OllamaProvider:
+    """Ollama provider implementation for the LLM protocol."""
+
+    __slots__ = (
+        "_host",
+        "_client",
+    )
+
+    name: ClassVar[str] = "ollama"
+
+    def __init__(
+        self,
+        /,
+        *,
+        host: str | None = None,
+    ) -> None:
+        self._host = _resolve_ollama_host(host)
+        self._client = Client(host=self._host)
+
+    def generate_text(
+        self,
+        /,
+        *,
+        model: str,
+        instructions: str,
+        user_text: str,
+    ) -> LLMTextResult:
+        """Generate text using Ollama (non-streaming)."""
+
+        messages = [
+            {"role": "system", "content": instructions},
+            {"role": "user", "content": user_text},
+        ]
+
+        try:
+            response = self._client.chat(model=model, messages=messages)
+        except ResponseError as exc:
+            raise RuntimeError(
+                f"Ollama API error: {exc}"
+            ) from exc
+        except Exception as exc:
+            raise RuntimeError(
+                f"Failed to connect to Ollama at {self._host}. Make sure Ollama is running: {exc}"
+            ) from exc
+
+        response_text = response.message.content or ""
+
+        # Extract token usage if available
+        prompt_tokens: int | None = None
+        completion_tokens: int | None = None
+        total_tokens: int | None = None
+
+        if hasattr(response, "prompt_eval_count") and response.prompt_eval_count:
+            prompt_tokens = response.prompt_eval_count
+        if hasattr(response, "eval_count") and response.eval_count:
+            completion_tokens = response.eval_count
+        if prompt_tokens is not None and completion_tokens is not None:
+            total_tokens = prompt_tokens + completion_tokens
+
+        return LLMTextResult(
+            text=response_text.strip(),
+            response_id=None,
+            usage=LLMUsage(
+                prompt_tokens=prompt_tokens,
+                completion_tokens=completion_tokens,
+                total_tokens=total_tokens,
+            ),
+        )
+
+    def count_tokens(
+        self,
+        /,
+        *,
+        model: str,
+        text: str,
+    ) -> int:
+        """Approximate token count using tiktoken; fallback to whitespace split."""
+
+        try:
+            encoding = _get_encoding()
+            return len(encoding.encode(text))
+        except Exception:
+            return len(text.split())

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: git-commit-message
-Version: 0.7.0
+Version: 0.8.0
 Summary: Generate Git commit messages from staged changes using LLM
 Maintainer-email: Mina Her <minacle@live.com>
 License: This is free and unencumbered software released into the public domain.
@@ -45,16 +45,22 @@ Requires-Python: >=3.13
 Description-Content-Type: text/markdown
 Requires-Dist: babel>=2.17.0
 Requires-Dist: google-genai>=1.56.0
+Requires-Dist: ollama>=0.4.0
 Requires-Dist: openai>=2.6.1
 Requires-Dist: tiktoken>=0.12.0
 
 # git-commit-message
 
-Staged changes -> GPT commit message generator.
+Generate a commit message from your staged changes using OpenAI, Google Gemini, or Ollama.
 
 [![asciicast](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN.svg)](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN)
 
-## Install (PyPI)
+## Requirements
+
+- Python 3.13+
+- A Git repo with staged changes (`git add ...`)
+
+## Install
 
 Install the latest released version from PyPI:
 
@@ -78,19 +84,43 @@ Quick check:
 git-commit-message --help
 ```
 
-Set your API key (POSIX sh):
+## Setup
+
+### OpenAI
 
 ```sh
 export OPENAI_API_KEY="sk-..."
 ```
 
-Or for the Google provider:
+### Google Gemini
 
 ```sh
 export GOOGLE_API_KEY="..."
 ```
 
-Note (fish): In fish, set it as follows.
+### Ollama (local models)
+
+1. Install Ollama: https://ollama.ai
+2. Start the server:
+
+```sh
+ollama serve
+```
+
+3. Pull a model:
+
+```sh
+ollama pull mistral
+```
+
+Optional: set defaults:
+
+```sh
+export GIT_COMMIT_MESSAGE_PROVIDER=ollama
+export OLLAMA_MODEL=mistral
+```
+
+Note (fish):
 
 ```fish
 set -x OPENAI_API_KEY "sk-..."
@@ -104,96 +134,116 @@ python -m pip install -e .
 
 ## Usage
 
-- Print commit message only:
+Generate and print a commit message:
 
 ```sh
 git add -A
 git-commit-message "optional extra context about the change"
 ```
 
-- Force single-line subject only:
+Generate a single-line subject only:
 
 ```sh
 git-commit-message --one-line "optional context"
 ```
 
-- Select provider (default: openai):
+Select provider:
 
 ```sh
-git-commit-message --provider openai "optional context"
+# OpenAI (default)
+git-commit-message --provider openai
+
+# Google Gemini (via google-genai)
+git-commit-message --provider google
+
+# Ollama
+git-commit-message --provider ollama
 ```
 
-- Select provider (Google Gemini via google-genai):
+Commit immediately (optionally open editor):
 
 ```sh
-git-commit-message --provider google "optional context"
+git-commit-message --commit "refactor parser for speed"
+git-commit-message --commit --edit "refactor parser for speed"
 ```
 
-- Limit subject length (default 72):
+Limit subject length:
 
 ```sh
-git-commit-message --one-line --max-length 50 "optional context"
+git-commit-message --one-line --max-length 50
 ```
 
-- Chunk long diffs by token budget (0 = single chunk + summary, -1 = disable chunking):
+Chunk/summarise long diffs by token budget:
 
 ```sh
 # force a single summary pass over the whole diff (default)
-git-commit-message --chunk-tokens 0 "optional context"
+git-commit-message --chunk-tokens 0
 
 # chunk the diff into ~4000-token pieces before summarising
-git-commit-message --chunk-tokens 4000 "optional context"
+git-commit-message --chunk-tokens 4000
 
 # disable summarisation and use the legacy one-shot prompt
-git-commit-message --chunk-tokens -1 "optional context"
+git-commit-message --chunk-tokens -1
 ```
 
-- Commit immediately with editor:
+Select output language/locale (IETF language tag):
 
 ```sh
-git-commit-message --commit --edit "refactor parser for speed"
+git-commit-message --language en-US
+git-commit-message --language ko-KR
+git-commit-message --language ja-JP
 ```
 
-- Print debug info (prompt/response + token usage):
+Print debug info:
 
 ```sh
-git-commit-message --debug "optional context"
+git-commit-message --debug
 ```
 
-- Select output language/locale (default: en-GB):
+Configure Ollama host (if running on a different machine):
 
 ```sh
-# American English
-git-commit-message --language en-US "optional context"
+git-commit-message --provider ollama --host http://192.168.1.100:11434
+```
 
-# Korean
-git-commit-message --language ko-KR
+## Options
 
-# Japanese
-git-commit-message --language ja-JP
-```
+- `--provider {openai,google,ollama}`: provider to use (default: `openai`)
+- `--model MODEL`: model override (provider-specific)
+- `--language TAG`: output language/locale (default: `en-GB`)
+- `--one-line`: output subject only
+- `--max-length N`: max subject length (default: 72)
+- `--chunk-tokens N`: token budget per diff chunk (`0` = single summary pass, `-1` disables summarisation)
+- `--debug`: print request/response details
+- `--commit`: run `git commit -m <message>`
+- `--edit`: with `--commit`, open editor for final message
+- `--host URL`: host URL for providers like Ollama (default: `http://localhost:11434`)
+
+## Environment variables
 
-Notes:
+Required:
 
-- The model is instructed to write using the selected language/locale.
-- In multi-line mode, the only allowed label ("Rationale:") is also translated into the target language.
+- `OPENAI_API_KEY`: when provider is `openai`
+- `GOOGLE_API_KEY`: when provider is `google`
 
-Environment:
+Optional:
 
-- `OPENAI_API_KEY`: required when provider is `openai`
-- `GOOGLE_API_KEY`: required when provider is `google`
-- `GIT_COMMIT_MESSAGE_PROVIDER`: optional (default: `openai`). `--provider` overrides this value.
-- `GIT_COMMIT_MESSAGE_MODEL`: optional model override (defaults: `openai` -> `gpt-5-mini`, `google` -> `gemini-2.5-flash`)
-- `OPENAI_MODEL`: optional OpenAI-only model override
-- `GIT_COMMIT_MESSAGE_LANGUAGE`: optional (default: `en-GB`)
-- `GIT_COMMIT_MESSAGE_CHUNK_TOKENS`: optional token budget per diff chunk (default: 0 = single chunk + summary; -1 disables summarisation)
+- `GIT_COMMIT_MESSAGE_PROVIDER`: default provider (`openai` by default). `--provider` overrides this.
+- `GIT_COMMIT_MESSAGE_MODEL`: model override for any provider. `--model` overrides this.
+- `OPENAI_MODEL`: OpenAI-only model override (used if `--model`/`GIT_COMMIT_MESSAGE_MODEL` are not set)
+- `OLLAMA_MODEL`: Ollama-only model override (used if `--model`/`GIT_COMMIT_MESSAGE_MODEL` are not set)
+- `OLLAMA_HOST`: Ollama server URL (default: `http://localhost:11434`)
+- `GIT_COMMIT_MESSAGE_LANGUAGE`: default language/locale (default: `en-GB`)
+- `GIT_COMMIT_MESSAGE_CHUNK_TOKENS`: default chunk token budget (default: `0`)
 
-Notes:
+Default models (if not overridden):
 
-- If token counting fails for your provider while chunking, try `--chunk-tokens 0` (default) or `--chunk-tokens -1`.
+- OpenAI: `gpt-5-mini`
+- Google: `gemini-2.5-flash`
+- Ollama: `gpt-oss:20b`
 
-## AI‑generated code notice
+## AI-generated code notice
 
 Parts of this project were created with assistance from AI tools (e.g. large language models).
-All AI‑assisted contributions were reviewed and adapted by maintainers before inclusion.
+All AI-assisted contributions were reviewed and adapted by maintainers before inclusion.
 If you need provenance for specific changes, please refer to the Git history and commit messages.

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message.egg-info/SOURCES.txt

@@ -8,6 +8,7 @@ src/git_commit_message/_gemini.py
 src/git_commit_message/_git.py
 src/git_commit_message/_gpt.py
 src/git_commit_message/_llm.py
+src/git_commit_message/_ollama.py
 src/git_commit_message.egg-info/PKG-INFO
 src/git_commit_message.egg-info/SOURCES.txt
 src/git_commit_message.egg-info/dependency_links.txt

{git_commit_message-0.7.0 → git_commit_message-0.8.0}/src/git_commit_message.egg-info/requires.txt

@@ -1,4 +1,5 @@
 babel>=2.17.0
 google-genai>=1.56.0
+ollama>=0.4.0
 openai>=2.6.1
 tiktoken>=0.12.0

git_commit_message-0.7.0/README.md

@@ -1,149 +0,0 @@
-# git-commit-message
-
-Staged changes -> GPT commit message generator.
-
-[![asciicast](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN.svg)](https://asciinema.org/a/jk0phFqNnc5vaCiIZEYBwZOyN)
-
-## Install (PyPI)
-
-Install the latest released version from PyPI:
-
-```sh
-# User environment (recommended)
-python -m pip install --user git-commit-message
-
-# Or system/virtualenv as appropriate
-python -m pip install git-commit-message
-
-# Or with pipx for isolated CLI installs
-pipx install git-commit-message
-
-# Upgrade to the newest version
-python -m pip install --upgrade git-commit-message
-```
-
-Quick check:
-
-```sh
-git-commit-message --help
-```
-
-Set your API key (POSIX sh):
-
-```sh
-export OPENAI_API_KEY="sk-..."
-```
-
-Or for the Google provider:
-
-```sh
-export GOOGLE_API_KEY="..."
-```
-
-Note (fish): In fish, set it as follows.
-
-```fish
-set -x OPENAI_API_KEY "sk-..."
-```
-
-## Install (editable)
-
-```sh
-python -m pip install -e .
-```
-
-## Usage
-
-- Print commit message only:
-
-```sh
-git add -A
-git-commit-message "optional extra context about the change"
-```
-
-- Force single-line subject only:
-
-```sh
-git-commit-message --one-line "optional context"
-```
-
-- Select provider (default: openai):
-
-```sh
-git-commit-message --provider openai "optional context"
-```
-
-- Select provider (Google Gemini via google-genai):
-
-```sh
-git-commit-message --provider google "optional context"
-```
-
-- Limit subject length (default 72):
-
-```sh
-git-commit-message --one-line --max-length 50 "optional context"
-```
-
-- Chunk long diffs by token budget (0 = single chunk + summary, -1 = disable chunking):
-
-```sh
-# force a single summary pass over the whole diff (default)
-git-commit-message --chunk-tokens 0 "optional context"
-
-# chunk the diff into ~4000-token pieces before summarising
-git-commit-message --chunk-tokens 4000 "optional context"
-
-# disable summarisation and use the legacy one-shot prompt
-git-commit-message --chunk-tokens -1 "optional context"
-```
-
-- Commit immediately with editor:
-
-```sh
-git-commit-message --commit --edit "refactor parser for speed"
-```
-
-- Print debug info (prompt/response + token usage):
-
-```sh
-git-commit-message --debug "optional context"
-```
-
-- Select output language/locale (default: en-GB):
-
-```sh
-# American English
-git-commit-message --language en-US "optional context"
-
-# Korean
-git-commit-message --language ko-KR
-
-# Japanese
-git-commit-message --language ja-JP
-```
-
-Notes:
-
-- The model is instructed to write using the selected language/locale.
-- In multi-line mode, the only allowed label ("Rationale:") is also translated into the target language.
-
-Environment:
-
-- `OPENAI_API_KEY`: required when provider is `openai`
-- `GOOGLE_API_KEY`: required when provider is `google`
-- `GIT_COMMIT_MESSAGE_PROVIDER`: optional (default: `openai`). `--provider` overrides this value.
-- `GIT_COMMIT_MESSAGE_MODEL`: optional model override (defaults: `openai` -> `gpt-5-mini`, `google` -> `gemini-2.5-flash`)
-- `OPENAI_MODEL`: optional OpenAI-only model override
-- `GIT_COMMIT_MESSAGE_LANGUAGE`: optional (default: `en-GB`)
-- `GIT_COMMIT_MESSAGE_CHUNK_TOKENS`: optional token budget per diff chunk (default: 0 = single chunk + summary; -1 disables summarisation)
-
-Notes:
-
-- If token counting fails for your provider while chunking, try `--chunk-tokens 0` (default) or `--chunk-tokens -1`.
-
-## AI‑generated code notice
-
-Parts of this project were created with assistance from AI tools (e.g. large language models).
-All AI‑assisted contributions were reviewed and adapted by maintainers before inclusion.
-If you need provenance for specific changes, please refer to the Git history and commit messages.